Monitoring Fundamentals

  • Article

Monitoring your cluster involves tracking data about the functional health, performance, and availability of your cluster and cluster members. Microsoft Application Center 2000 (Application Center) simplifies this process by gathering, logging, and presenting this information in charts and lists. Application Center also includes tools that you can use to monitor your cluster and cluster members and to automate notifying the administrator and taking corrective actions on a cluster. You can monitor clusters and members with several goals in mind, such as monitoring Web site availability, determining if you need to add or subtract members, planning for future usage spikes, and locating performance bottlenecks.

Application Center continuously records and displays four aspects of the functional health of your cluster and members: events, performance counters, server health (by using Microsoft Health Monitor 2.1 data collectors), and member status. This data informs you of the availability of your cluster and members and of problems that might be developing.

Event and monitoring data is generated by various services, applications, and objects, including the following:

  • Microsoft® Windows® 2000 events, which you can use to determine the functional status of Application Center and Windows 2000.

  • Microsoft Windows Management Instrumentation (WMI), which provides a comprehensive and extensible management infrastructure for applications, services, and objects in Windows 2000. It provides a standardized method for accessing events and other monitoring and management information across an enterprise environment.

  • Health Monitor, which sends events and executes actions in response to exceeded thresholds by using data collectors. Application Center provides several Health Monitor data collectors, which are configured by default. You can add to these by using Health Monitor to create new data collectors or by installing the Health Monitor sample data collectors.

  • Application Center, which is the most frequent supplier of events to itself. Unlike other event providers, Application Center also includes extended information about its events.

Application Center Monitoring Functionality

Application Center offers the following core monitoring functionality:

  • Events, which are occurrences on your cluster and members that are of administrative interest. Events are generated by a range of objects, services, and applications within Application Center and throughout your network and are displayed in the Application Center Events view. When Application Center Event and Performance Logging is installed, error events are logged to the Windows Event Log; when Application Center Event and Performance Logging is not installed, both Warning and Error events are logged to the Application Center log. You can use the Windows Event Log to view any Application Center event that is logged to the Windows Event Log.

  • Health Monitor data collectors, which are objects that collect metrics for a particular process or service and use thresholds. Thresholds compare a particular metric in the data collector to a user-defined limit or condition. When that threshold is met, the threshold triggers a status change in the data collector which triggers an action, such as sending an e-mail alert to the system administrator.

  • Performance counters, which indicate how much a specific resource is used, such as a certain percentage of the hard disk. Application Center logs several default performance counters to give you an overview of the performance for your cluster and members. By using the Application Center Performance view, you can see the counters in real time or for a set of pre-defined periods. You can also use the Windows 2000 Performance Viewer.

    By default, no performance counters are displayed in the user interface. To view counters, you must add them. Additional sample performance counters are available for installation.

Monitoring Tasks

Most frequently, you will use monitoring after you have set up your cluster to ensure that your content and services are at their highest availability and performance. This involves watching events that signify that a member or Web site has been taken offline, as well as watching performance counters for situations such as memory leaks and CPU resource use.

Additional monitoring tasks include the following:

  • Analyzing hardware performance by monitoring CPU use, memory consumption, file serving performance, and disk space.

  • Analyzing application efficiency by monitoring performance and resource use.

  • Analyzing network availability by monitoring for situations where memory and CPU resources are normal, but requests are not being serviced, or when users are not able to make a connection to your cluster.

  • Creating and editing data groups, data collectors, thresholds, and actions.

  • Analyzing statistics gathered from data collectors.

  • Executing automatic actions in response to status changes.

  • For more information about Windows 2000 events, see the Windows 2000 Help.

  • For more information about the Application Center monitoring process, see Monitoring Process and Available Data.

  • For more information about performance counters, see Performance Counters.

  • For more information about Health Monitor data collectors, see Using Monitoring and the Health Monitor Help.

  • For more information about monitoring tasks, see How To....

Did you find this information useful? Please send your suggestions and comments about the documentation to acdocs@microsoft.com.